Journal
Two pieces a week about using artificial intelligence in real work: which process is worth automating, which model to choose, what an agent actually changes, and where AI fails. Written for lawyers, accountants, marketing leads and operators — not for developers.
Territories
Which part of my process can AI actually automate today?
Which AI model should I use for each kind of work?
What does an AI agent do that a chat cannot?
Why does AI fail so confidently — and what do you do about it?
What can I put inside an AI tool without creating a legal problem?
Published
A language model does not look anything up. It predicts the next piece of text that best fits everything before it, so an invented citation comes out of the same mechanism that produces a correct one. The model has no internal signal separating the two. That is why the fix is never to ask it to be more careful. The fix is to hand it the source and check the answer against that source.
4 min read
